Theor Appl Genet (2011) 122:1029–1037 DOI 10.1007/s00122-010-1507-2
ORIGINAL PAPER
Linkage map construction involving a reciprocal translocation A. Farre´ • I. Lacasa Benito • L. Cistue´ • J. H. de Jong • I. Romagosa • J. Jansen
Received: 19 August 2010 / Accepted: 25 November 2010 / Published online: 14 December 2010 Ó The Author(s) 2010. This article is published with open access at Springerlink.com
Abstract This paper is concerned with a novel statistical–genetic approach for the construction of linkage maps in populations obtained from reciprocal translocation heterozygotes of barley (Hordeum vulgare L.). Using standard linkage analysis, translocations usually lead to ‘pseudolinkage’: the mixing up of markers from the chromosomes involved in the translocation into a single linkage group. Close to the translocation breakpoints recombination is severely suppressed and, as a consequence, ordering markers in those regions is not feasible. The novel strategy presented in this paper is based on (1) disentangling the ‘‘pseudo-linkage’’ using principal coordinate analysis, (2) separating individuals into translocated types and normal types and (3) separating markers into those close to and those more distant from the translocation breakpoints. The methods make use of a consensus map of the species involved. The final product consists of integrated linkage maps of the distal parts of the chromosomes involved in the translocation.
Communicated by C. Hackett. A. Farre´ I. Lacasa Benito I. Romagosa Department of Plant Production and Forest Science, University of Lleida, Lleida, Spain L. Cistue´ Estacio´n Experimental de Aula Dei, CSIC, Zaragoza, Spain J. H. de Jong Laboratory of Genetics, Wageningen University and Research Centre, Wageningen, The Netherlands A. Farre´ J. Jansen (&) Biometris, Wageningen University and Research Centre, Wageningen, The Netherlands e-mail:
[email protected]
Introduction At present barley (Hordeum vulgare L.) is the fourth most important cereal crop, after maize, wheat and rice. Barley is a diploid (2n = 14) self-pollinating species which can be used as a model plant for other Triticeae species providing an excellent system for genome mapping and map-based analyses (Costa et al. 2001). Well-established linkage maps containing morphological, chromosomal, protein and molecular markers are indispensable tools for genetic analysis and for analysing DNA sequence variation between the crops and its relative accessions and species. In barley, common molecular marker technologies include restriction digest based RFLPs (Restriction Fragment Length Polymorphisms; Graner et al. 1991; Heun et al. 1991; Kleinhofs et al. 1993), and PCR-based markers like AFLPs (Amplified Fragment Length Polymorphism; Waugh et al. 1997), STS (Sequence Tagged Site; Olson et al. 1989), SSRs (Simple Sequence Repeat; Liu et al. 1996; Ramsay et al. 2000; Pillen et al. 2000; Li et al. 2003), and recently, DArT (Diversity Arrays Technology), a technique based on DNA hybridisation (Jaccoud et al. 2001; Wenzl et al. 2004). The more recent SNPs (Single Nucleotide Polymorphisms) have become popular thanks to a high density within the genome, easy development from sequence data and high reproducibility. Wenzl et al. (2006) constructed a high-density linkage map which will be used as a reference for future mapping work. The map was based on datasets from multiple populations, comprising 2,935 loci (2,085 DArT, 850 SSR, RFLP and/or STS markers) and spans 1,161 cM. Recently, Varshney et al. (2007) produced a high-density linkage map, using SSR marker data from six mapping populations. A total of 775 microsatellite loci from 688 primer pairs were used to cover 1,068 cM of the barley genome.
123
1030
Major complications in the construction of genetic maps arose in populations containing reciprocal translocation heterozygotes (Lehmensiek et al. 2005). Reciprocal translocations involve the exchange of two terminal segments between two non-homologous chromosomes. Due to reciprocal translocations recombination between loci around the translocation breakpoints will be suppressed. Consequently, markers in these regions become ‘pseudo-linked’, i.e. linkage is detected between markers lying on different chromosomes. Reciprocal translocations are induced spontaneously or artificially by ionizing radiation, mutagens or by transposons and can be confirmed by aberrant chromosome and banding morphology, by establishing quadrivalent formation in metaphase I complements, and semi-sterility of pollen in translocation heterozygotes, and by linkage analysis (Ja´uregui et al. 2001). Translocations are well documented in various crops, including rye (Alonso-Blanco et al. 1993; Benito et al. 1994; Catarino et al. 2006), soybean (Mahama and Palmer 2003), Prunus (Ja´uregui et al. 2001), Lens (Tadmor et al. 1987) and pea (Kosterin et al. 1999). Many bread wheat cultivars contain the 1RS-1BL wheat–rye chromosome translocation, introduced to provide resistance to pests and diseases (Mago et al. 2002; Sharma et al. 2009). In barley, spontaneous reciprocal translocations were described by Konishi and Linde-Laursen (1988) who investigated 1,240 barley landraces and varieties and 120 wild barley lines, using test-cross and Giemsa chromosome banding analyses. Of the 1,240 landraces and varieties, four carried translocations involving chromosomes 2H and 4H with breakpoints near or at the centromere. Of the 120 wild barley lines, three carried translocations between chromosomes 2H and 4H, 3H and 5H, 3H and 6H, respectively. ‘Albacete’ is a barley variety that it is known to have a reciprocal chromosomal translocation. Recent results suggest that chromosomes 1H and 3H are involved in this translocation (Lacasa-Benito et al. 2005). The consensus map of barley (Wenzl et al. 2006) was used as a reference. Preliminary linkage analysis of a doubled haploid population obtained from a cross between ‘Albacete’ and the non-translocation barley variety Barberousse shows that linkage groups 1H and 3H are indeed involved in the translocation. In order to detect unexpected relationships between markers at those chromosomes we apply a principal coordinate analysis which visualises marker positions in a three-dimensional graph. The next step of our analysis is based on the cytogenetical theory that (1) during meioses two types of viable gametes are produced: a ‘‘normal’’ type (with a balanced set of nontranslocation chromosomes) and a ‘‘translocated’’ type (with a balanced set of translocated chromosomes), and (2) that around the translocation breakpoints recombination is
123
Theor Appl Genet (2011) 122:1029–1037
suppressed (Burnham 1962; Sybenga 1975). The crucial step consists of dividing the doubled haploid population into two subpopulations according to the origin of the alleles at the translocation breakpoints: Barberousse alleles refer to the ‘‘normal’’ type and Albacete alleles refer to the ‘‘translocated’’ type. The position of the translocation breakpoint is found by a graphical display of recombination frequencies versus positions on the consensus map. At this stage we discuss the problem of extreme segregation distortion around the translocation breakpoints in the two subpopulations, which is solved by removing markers around the translocation breakpoints with severe segregation distortion. Linkage maps for the various chromosome arms are then obtained for each of the subpopulations separately. The final step is integrating the linkage maps of the corresponding arms of linkage groups 1H and 3H of the two subpopulations.
Materials and methods Plant material Two 6-row parental barley varieties, ‘Albacete’ and ‘Barberousse’, were crossed and anther culture lines were produced from the F1 from which 231 spontaneous doubled haploids were obtained at Estacio´n Experimental de Aula Dei, Zaragoza, Spain. Seed from each doubled haploid line was retained as a reference stock and used to grow the plants for DNA extraction. More than 50 years ago, ‘Albacete’ was found by Enrique Sa´nchez-Monge at the Estacio´n Experimental Aula Dei, Spanish Research Council, as an inbred line in a local heterogeneous landrace population from Albacete (province of Spain). It is a sixrow barley variety, drought tolerant with a stable grain yield. For these properties, it has been grown for decades on up to 1 million ha/year, especially in drought-prone areas. ‘Barberousse’ is a six-row winter barley variety released in France in 1977. It was obtained from the cross [259 711/(Hatif de Grignon/Ares)]/Ager. Although it is sensitive to drought, this variety is known for its good productivity and easy adaptation. DNA isolation Genomic DNA of the DHs was extracted using two different protocols. For genomic and EST-derived SSR markers, genomic DNA was extracted from 0.1 g of fresh leaf tissue sprayed with liquid nitrogen according to the method of Doyle and Doyle (1990) with some modifications. DNA was diluted in TE buffer pH8 and preserved for a long period at a temperature of -20°C. For DArT
Theor Appl Genet (2011) 122:1029–1037
markers the genomic DNA was extracted from fresh leaf tissue using the Qiagen DNeasy 96 Plant Kit.
1031
Results Polymorphisms between parents
Genotyping Genotypic data were first produced for genomic and ESTderived SSR markers. We refer to the Genomic SSRs markers using the prefix (‘‘Bmac’’ or ‘‘Bmag’’) followed by four digits and to the EST-derived SSR markers using the prefix (‘‘scssr’’) followed by five digits. Genomic and ESTderived SSR analyses were carried out using the fluorescent fragment detection system on an ABI PRISMÒ 3100-Avant Genetic Analyzer using Genescan software from Applied Biosystems. For this method, the reverse primer was labelled with a fluorochrome at the 50 end. The PCR programmes used for all amplifications are published in the UK CropNet databases (http://ukcrop.net/db.html). Subsequently, a genome-wide scan was carried out using the Diversity Arrays Technology (DArT, http://www. diversityarrays.com). In a single assay, DArT uses microarrays to detect DNA polymorphisms at several hundred genomic loci spread over the entire genome. It is based on hybridising labelled genomic representations of individual DNA samples on a micro-array, which contains a large number of DNA fragments derived from the total genomic DNA of the species under consideration (Wenzl et al. 2004). Polymorphisms were scored as presence (=1) or absence (=0) of hybridisation to individual array elements. The locus designations of Triticarte Pty. Ltd. were used. DArT markers consisted of prefix (‘‘bPb’’ or ‘‘CloneID’’) followed by a number corresponding to the particular clone in the genomic representation. Linkage analysis For the construction of linkage maps we used the maximum likelihood (ML) algorithm in Jansen et al. (2001) which enables map construction in batch mode and the implementation in JoinMap 4.0 (van Ooijen 2006) which enables the use of many graphical tools. Linkage maps were drawn using MapChart 2.2 (Voorrips 2002). Statistical techniques Principal coordinate analysis (Chatfield and Collins 1980), a special form of multidimensional scaling, was used to obtain a 3D representation of the markers on the linkage groups involved in the reciprocal translocation. The similarity between markers was measured by the simple matching coefficient, which for a doubled haploid population is equal to one minus the recombination frequency. Calculations were carried with Genstat version 12 (Payne et al. 2009), using the directives FSIMILARITY and PCO.
We applied the following liberal rules for excluding markers: (1) markers with an unknown location on the consensus map (these were later included again when constructing our final maps); (2) markers with an ambiguous chromosome assignment on the consensus map; (3) markers with more than 5% missing data and (4) markers with low quality readings (Q value less than 77). For linkage analysis, 30 polymorphic genomic and ESTderived SSR markers and 309 polymorphic DArT were used. This amounted to an average of 48 markers per chromosome, with an average interlocus distance of 4.4 cM based on the consensus linkage map. A few regions had gaps of over 10 cM. Chromosome 4H showed strikingly low levels of polymorphisms, which is in agreement with the findings of Qi et al. (1996), Karakousis et al. (2003), Kleinhofs et al. (1993) and Wenzl et al. (2006). Preliminary linkage map construction We obtained six main linkage groups involving 305 markers. Using a recombination frequency threshold of 0.2, 34 markers could not be assigned to one of these linkage groups. Five linkage groups could be related to linkage groups of the consensus map (Wenzl et al. 2006): 2H (48 markers), 4H (20 markers), 5H (48 markers), 6H (49 markers) and 7H (58 markers). The remaining 82 markers had known positions on 1H and 3H of the consensus map, of which 78 formed one linkage group. Four markers, on the consensus map located at the very end of the long arm of linkage group 1H, were found to be unlinked with the 78 markers of 1H and 3H. One marker showing a consistently high nearest-neighbour stress (Van Ooijen 2006) was also excluded. Figure 1 displays one linkage map with the positions of the remaining 77 markers obtained by the maximum likelihood algorithm. Many different linkage maps were obtained by rerunning the maximum likelihood algorithm. This was a clear indication that it was not possible to obtain a linear order of the markers in an unambiguous way. A spatial representation of the positions of the markers was obtained using principal coordinate analysis. Figure 2 shows a three-dimensional plot in which the points represent the positions of the markers with regard to the first three principal axes; the percentage of variance explained by the first three axes was 27.2, 15.5 and 13.9, respectively. The points have been joined by straight lines according to their position on the consensus map of Wenzl et al. (2004). Markers of linkage group 1H and 3H have been displayed in blue and ochre, respectively. Figure 2 shows that linkage
123
1032
Theor Appl Genet (2011) 122:1029–1037 bPb-1165 bPb-3622 bPb-1348 bPb-3776 bPb-8973 bPb-9414 bPb-7306 bPb-9337 bPb-7231 bPb-6408 bPb-1398 bPb-1535 bPb-1231 bPb-1586 bPb-9423 bPb-9333 bPb-429 bPb-4949 bPb-9717 bPb-6621 bPb-910 Bmag718 Bmac0209 Bmag0006Bmac0067 EBmac0501 bPb-9360 bPb-1193 bPb-1791 bPb-1150 bPb-1723 bPb-9005 bPb-3389 bPb-3089 bPb-4144 bPb-1213 bPb-5249 bPb-9108 bPb-5014
0 50 100 150 200 250
bPb-1681 bPb-6765 bPb-3317 bPb-8410 bPb-40 bPb-3805 bPb-94 bPb-5771 bPb-5012 bPb-8321 Bmag0136 HvM20 bPb-9746 bPb-2929 bPb-7448 bPb-7199 bPb-7770 bPb-9583 bPb-9945 bPb-1264 bPb-4022 bPb-5555
300 350 400 450
bPb-8506 bPb-2550 bPb-8621 bPb-9111 bPb-5666 bPb-3899 bPb-9599 bPb-789 bPb-6383 bPb-7827 bPb-7815 bPb-8419 bPb-200 bPb-361 bPb-7724 scssr25538
500 550
Fig. 1 Graphical representation of a linkage map obtained for markers of linkage groups 1H and 3H according to the consensus map of Wenzl et al. (2006); markers of linkage group 1H and 3H are displayed in blue/dark grey and ochre/light grey, respectively
3HS
3HL
1HS
is 3rd ax
1HL
2 nd axis
is 1st ax
Fig. 2 Graphical representation of a principal coordinate analysis using the simple matching coefficient for the markers of linkage groups 1H and 3H according to the consensus map of Wenzl et al. (2006). The percentage of variance explained by the first three principal axes is 27.2, 15.5 and 13.9, respectively. The markers have been joined according to their position on the consensus map of Wenzl et al. (2006). The black circle encloses markers of linkage groups 1H and 3H that are close to the translocation breakpoints; they are separated by a small number of recombinations
groups 1H and 3H merge at one point (indicated by the circle) and that the short and long chromosome arms divert from this point, and from each other, in three-dimensional space. As a consequence, the translocation breakpoints lie within the circle.
123
Recombination around the translocation breakpoints To study the recombination behaviour of markers around the translocation breakpoints
